A PhD candidate writes a CQL library to screen patients for a Hospital at Home
program. It runs at 30 to 40 seconds per patient. He needs it to run across
300,000. He has no idea whether the problem is his logic, his engine, or his data.
He gets lucky. His email reaches someone who has spent years optimizing CQL, and
that person writes back a detailed review: functions are not cached the way defines
are, a retrieve iterates every resource in the bundle every time, this correlated
query is really a lateral join running dozens of semi‑joins. Rewrite it this way.
It worked. He restructured the library around prefiltered, cached defines. The same
logic now runs in under a second and returns the same results, down from 30 to 40.
Four people read that answer. It is not searchable, it is not attached to any
measure, and the next person with the same problem will not find it. It existed
because somebody made an introduction.
This is a great example of why we need an open community for quality where experts
can share their expertise, rather than relying on an introduction to the one person
who happened to know someone by pure luck.
— from the thread that started this
Every payer, health system, and vendor doing quality measurement staffs its own team
to implement the same national measures against its own data, then pays auditors to
confirm the implementation is right. The same specification ambiguities get
rediscovered independently, over and over. The cost is enormous and none of it
reaches a patient.
Open Quality is where that answer should live.